The development of effective disease-modifying treatments for Alzheimer’s disease (AD) is constrained by the capacity to screen the right target population for clinical trials, broadly and cost-effectively. Both amyloid-β (Aβ) and tau Positron Emission Tomography (PET) scans as well as cerebrospinal fluid (CSF) analysis of surrogate biomarkers of AD provide prognostic ability, although they are too expensive or invasive to enable screening for AD in large preclinical populations. Blood-based biomarkers have achieved significant progresses due to the development of novel methodologies, such as the single molecule array (SiMoA) technology, an ultra-sensitive ELISA technique. The predictive capacity of these blood-based biomarkers decreases, however, significantly as a function of time prior to the onset of cognitive dysfunction. NeoNeuro has developed a novel, alternative approach – in collaboration with the INSIGHT-preAD study– involving the application of enriched aptamer libraries as a means of mapping multiple pathological epitopes in blood, both rapidly and cost-effectively.
Based on the cross-validation analysis, our Aptamarker platform has the following predictive capacity:
Sensitivity = 0.80
Specificity = 0.88
Accuracy = 0.83
AUC = 0.9694 in three dimensions
Figure 1: Distribution of 42 individual samples by dimensional contrast based on aptamer frequencies.
Figure 2 provides the area predicted for each of the classes of AB status considered, with the locations of each of the 42 samples within these prediction areas. These predicted areas are not subject to the cross-validation analysis, only the first two dimensions are illustrated.
Figure 3 provides the ROC curve following cross-validation analysis with 100 replications and 10 folds. The calculations are based on multivariate prediction methods with a threshold based on distance.The values provided are based on analysis across three principal components.